search text   Advanced Search

This Article Abstract Full Text (PDF) An erratum has been published P3Rs: Submit a response Alert me when this article is cited Alert me when P3Rs are posted Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Keren, R. Articles by Coffin, S. E. Search for Related Content PubMed PubMed Citation Articles by Keren, R. Articles by Coffin, S. E. Related Collections Infectious Disease & Immunity

Direct Medical Cost of Influenza-Related Hospitalizations in Children

^a General Pediatrics
^b Infectious Diseases
^c Biostatistics and Epidemiology, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 
OBJECTIVE. Our goal was to determine the cost of influenza-related hospitalization in children with community-acquired laboratory-confirmed influenza and to identify predictors of high hospitalization costs.

PATIENTS AND METHODS. This was a retrospective cohort study of patients 21 years and younger hospitalized at a children's hospital with community-acquired laboratory-confirmed influenza during 4 consecutive influenza seasons (2000–2004). The main outcome measure was the direct medical cost of influenza-related hospitalizations, including the cost of diagnostics, therapeutics, room, and physician services.

RESULTS. Electronic billing data were retrievable for 727 (98%) of 745 patients hospitalized for community-acquired laboratory-confirmed influenza during the study period. A total of 478 (66%) children were in a high-risk group for whom the Advisory Committee on Immunization Practices recommended influenza vaccine (patients with Advisory Committee on Immunization Practices–designated chronic medical conditions or aged 6–23 months). The mean total cost of hospitalization for influenza-related illness was $13159 ($39792 for patients admitted to an ICU; $7030 for patients cared for exclusively on the wards). High-risk patients had higher mean total costs ($15269) than low-risk patients ($9107). Cardiac, metabolic, and neurologic/neuromuscular diseases and age of 18 to 21 years were independently associated with the highest hospitalization costs (>15th percentile).

CONCLUSIONS. The cost of influenza-related hospitalizations in children may be considerably higher than previously estimated. The presence of certain Advisory Committee on Immunization Practices–designated chronic medical conditions is associated with higher influenza-related hospitalization costs. Successfully immunizing patients with these conditions has the potential for significant cost savings.

Key Words: influenza • vaccines • cost analysis • hospitalization • children

Abbreviations: ICD-9—International Classification of Diseases, Ninth Revision • CA-LCI—community-acquired laboratory-confirmed influenza • CHOP—Children's Hospital of Philadelphia • RSV—respiratory syncytial virus • ACIP—Advisory Committee on Immunization Practices • OR—odds ratio • IQR—interquartile range

Influenza virus infections are responsible for 36000 deaths and 200000 hospitalizations in the United States each year.^1,2 Influenza-related deaths in children are rare, but influenza-attributable hospitalization rates in children younger than 2 years approach or, in the case of high-risk children, exceed hospitalization rates seen in adults 65 years and older.^1,3–7 Understanding the cost of these hospitalizations is important in evaluating the economic impact of influenza vaccination. Most studies have estimated the cost of a pediatric hospitalization for influenza at between $3000 and $4000.^8–10 However, these estimates were derived from administrative data for children hospitalized with diseases attributable to influenza (eg, International Classification of Diseases, Ninth Revision [ICD-9] codes 480–487 for pneumonia and influenza), rather than actual laboratory-confirmed cases. Thus, any misclassification of disease status could have resulted in inaccurate estimates of the cost of influenza-related hospitalizations.

In this study we analyzed the direct medical costs of hospitalization for a large cohort of children hospitalized with community-acquired laboratory-confirmed influenza (CA-LCI). Using detailed hospital billing data we sought to fully characterize the resources utilized in caring for children hospitalized with influenza and to identify factors associated with higher inpatient costs.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 
Design, Setting, and Patients
As part of a retrospective cohort study of patients hospitalized with CA-LCI at the Children's Hospital of Philadelphia (CHOP), we performed a cost analysis using electronic billing data. CHOP is an academic, tertiary care hospital with 418 patient beds and 22000 hospital admissions each year. We included patients 21 years and younger who were hospitalized with CA-LCI infection during 4 consecutive influenza seasons (June 2000 through May 2004). We have previously described the methods used for identifying cases of CA-LCI during the study period.¹¹ Briefly, we identified cases using both clinical virology laboratory records and hospital administrative data (influenza-specific ICD-9 admission or discharge codes 487.0, 487.1, and 487.8). Patients identified by ICD-9 codes required laboratory confirmation to be included as cases in the study cohort. Our ability to accurately identify most cases of influenza was augmented by CHOP's policy of testing all patients (for cohorting reasons) hospitalized with acute respiratory symptoms of unclear etiology for respiratory viral pathogens. Nasal-wash specimens were tested initially by immunoassay for respiratory syncytial virus (RSV) (Binax, Portland, ME) and influenza (Binax). Direct fluorescent antibody testing for adenovirus, influenza A and B, parainfluenza virus types 1, 2, and 3, and RSV was performed on specimens that tested negative by immunoassay for RSV and influenza. Finally, comprehensive viral culture was established for all specimens that were negative for respiratory viruses on the direct fluorescent antibody test.

Data Sources
We queried the hospital's enterprise data warehouse to obtain hospital billing data for each patient's influenza-related hospitalization. The hospital bills contain detailed line-item charges for all diagnostic tests, therapeutics, supplies, and room fees. Each charged item is dated and includes a unique service code, a description of the good or service, the billing department to which it belonged, the quantity provided, and the associated charge. The hospital's charge master index organizes charges into specific hospital billing departments. We used these billing departments to create broad categories of costs (diagnostics, therapeutics, and supplies/room) and more detailed subcategories (eg, pharmaceuticals, blood products, and intensive care therapies within the main category of therapeutics). The hospital bills contain daily ward-specific room charges (include charges for room, board, and nursing), which allowed us to distinguish ICU from non-ICU (ie, ward) days. Detailed chart review also provided us with information about patient demographics and the presence of Advisory Committee on Immunization Practices (ACIP)–designated high-risk characteristics¹² such as age 6–23 months and chronic medical conditions such as asthma, chronic lung diseases, cardiac disease, immunosuppression, hemoglobinopathies, chronic renal dysfunction, diabetes mellitus and inborn errors of metabolism, long-term salicylate therapy, neurologic and neuromuscular disease, and pregnancy. We did not have complete or reliable information on influenza vaccine status before hospitalization. For patients with multiple hospitalizations during the 4-year study period, we included only the first hospitalization in the analysis.

We also analyzed physician billing data from a separate database to estimate the cost of physician services provided during the hospitalization, such as daily visits, consultations, and procedures.

Analysis
We used the hospital's department-specific cost-to-charge ratios to convert the charges from the hospital billing data to costs. We used the National Physician Fee Schedule Relative Value File¹³ (calendar year 2004) to estimate the cost of physician services documented by Current Procedural Terminology codes in the physician billing data. Relative value units were adjusted for the geographic practice cost index for metropolitan Philadelphia, and the 2004 conversion factor of $37.3374 per relative value unit was used to calculate costs for each physician service. The cost of anesthesia services was estimated by using the Medicaid reimbursement rate of $18.61 per 15-minute time interval in the operating room or sedation suites. We used the medical component of the consumer price index¹⁴ to inflate all costs over the 4-year study period to 2004 US dollars. To validate the cost estimates derived by these methods, we compared them to reimbursements (adjusted for inflation).

We calculated mean costs (total and according to categories and subcategories) and compared costs of patients hospitalized for any duration in one of the ICUs (pediatric, neonatal, cardiac) and those hospitalized exclusively on the wards. We also compared costs for patients with and without ACIP-designated high-risk characteristics. The hospital costs were highly skewed; thus, they were log-transformed for all subgroup comparisons. We compared the means of the log-transformed costs using a z-score method, which is more appropriate than a t test when comparing means for 2 samples with unequal variance, as was the case for our data.^15,16 We used the Wilcoxon rank-sum test to compare length of stay (days) in subgroup analyses.

We developed a model to identify patient characteristics associated with high-cost hospitalizations, defined as total hospitalization costs of more than $15000, which corresponded to the 85th percentile of costs. We selected $15000 as the cutoff by visually identifying the beginning of the right tail of the distribution of total hospitalization costs. The independent variables included the chronic medical conditions outlined above and age. We summarized the independent variables, which were all defined as categorical, using frequencies and percentages. We used univariate and multivariable logistic regression to estimate odds ratios (ORs) for the association between the independent variables and high-cost hospitalizations.

All statistical calculations were performed using SAS 9.1 (SAS Institute Inc, Cary, NC). The institutional review board at CHOP approved the study.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 
Patient Characteristics
Over the 4-year study period we identified 745 patients hospitalized for CA-LCI. Electronic billing data were retrievable for 727 (98%) of these patients. There were slightly more boys than girls, and most (77%) of the children were under 5 years of age (Table 1). Fifty-five percent of the children were black, and 4% were Hispanic. A total of 349 (48%) patients had 1 or more chronic medical conditions recognized by the ACIP as risk factors for severe influenza infection. The most prevalent high-risk condition was asthma (25%). In addition, 129 previously healthy children aged 6 to 23 months were hospitalized with CA-LCI. Thus, a total of 478 (66%) children hospitalized with CA-LCI were in a group for whom the ACIP currently recommends influenza vaccine.

Risk Factors for High Resource Utilization
In bivariate analyses all the ACIP-designated high-risk conditions except immunosuppressive disorders and hemoglobinopathies were associated with total hospital costs greater than $15000 (Table 3). With the exception of children 5 to 11 years old, all other age groups were more likely than children 0 to 5 months old to have high-cost hospitalizations, although patients 18 to 21 years of age had the highest odds of having a high-cost hospitalization. In a multivariable model that included age (0–17 vs 18–21 years) and chronic conditions associated with high-cost hospitalization (at the P < .05 level in bivariate logistic-regression analyses), only cardiac, metabolic, and neurologic/neuromuscular diseases and age of 18 to 21 years were associated with high hospitalization costs.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

The mean cost of hospitalization for a child hospitalized with influenza was significantly greater than reported in most previous studies. In an evaluation of the economic impact of pandemic influenza, Meltzer et al¹⁰ estimated the direct medical costs of hospitalization using administrative data from the Marketscan Database, a proprietary database containing health insurance claims data from 4 million insured persons. On the basis of hospitalizations in which the discharge diagnosis included an ICD-9 code for a condition attributable to influenza (pneumonia and bronchitis [480–487.8], acute bronchitis [466–466.1], chronic respiratory disease [490–496], and heart-related conditions recorded as either the principal or a secondary diagnosis on the inpatient claim), they calculated a weighted mean cost of $4129 ± $2951 (2004 US dollars) for an influenza-related hospitalization in patients 0 to 19 years of age. In a cost-effectiveness analysis of intranasal influenza vaccine, Luce et al⁹ used the 1995 Nationwide Inpatient Sample to estimate the cost of an influenza-related hospitalization in the pediatric population. On the basis of hospitalizations in which discharge diagnoses included ICD-9 codes for pneumonia and bronchitis (480–487.8), they estimated a mean cost per influenza-related hospitalization of $2964 (2004 US dollars) and a mean length of stay of 3 days. Other authors have taken an even more indirect approach to estimating costs of hospitalization for influenza. In an economic evaluation of influenza vaccination in preschool children, Cohen and Nettleman⁸ averaged the mean length of stay reported in the National Hospital Discharge Survey¹⁷ for respiratory infection and pneumonia (3.4 days) and multiplied it by the average cost of a hospital day for respiratory disease reported in another study,¹⁸ resulting in an estimated cost of $3643 (2004 US dollars). The hospitalization costs reported in our study were closest in magnitude to those reported by Hall and Katz,¹⁹ which, interestingly, is the only other published study of children hospitalized with LCI. In that smaller study of 35 patients, the weighted mean hospitalization cost for children with and without ACIP-designated high-risk conditions was $19117 and $6072, respectively.

The hospitalization cost estimates of the first 3 previous studies were limited by their case-finding methods, which relied exclusively on ICD-9 codes that are not specific to influenza (eg, all the pneumonia and bronchitis codes [480–487.8]). The higher mean hospitalization costs documented in our study and the one by Hall and Katz¹⁹ may reflect the fact that true influenza-related hospitalizations generate greater resource utilization than hospitalizations for non–influenza-related pneumonia and bronchitis. Analyses of other administrative data sources support this notion. In 32 children's hospitals that contributed administrative data to the Child Health Corporation of America's Pediatric Health Information System, the mean total hospitalization cost for 1245 children discharged in 2004 with an influenza-specific ICD-9 code (487, 487.0, 487.1, and 487.8) as the primary or a secondary diagnosis was $21610 (IQR: $2184–$13238), compared with a mean total hospitalization cost of $5687 (IQR: $1512–$4609) for the 33681 children admitted with one of the pneumonia and bronchitis codes (480–487.8) as the primary diagnosis (unpublished data).

The higher mean costs found in our study could be attributed to inadequate adjustment of charges to cost by our hospital's cost accounting system. However, comparison of our calculated mean total cost to mean total charge and reimbursement suggests that our cost estimates are probably fairly accurate. The calculated mean total hospitalization cost ($13159 [IQR: $3896–$10882]) was approximately half the mean total charge ($26172 [IQR: $5805–$17196]), suggesting that the cost-to-charge adjustments were adequate. Our calculated mean total hospitalization cost was also very similar to mean total reimbursement ($14770 [IQR: $3190–$10105]), which some analysts use as another estimate of costs.

Case mix at our hospital also may have played a role in determining mean total hospitalization costs in our study. CHOP is a tertiary care center that admits a high proportion of children with complex chronic diseases, as evidenced by the fact that 349 (48%) of our patients had an ACIP-designated high-risk medical condition. A direct correlation between proportion of patients with high-risk conditions and estimates of influenza-related hospitalization costs can be seen in the Meltzer et al¹⁰ study (15% of patients 0–17 were assumed to have a high-risk condition) and the Hall et al cost analysis¹⁹ (57% with a high-risk condition), suggesting that case mix influences cost estimates in these analyses. CHOP also serves children from a large inner-city population that, because of problems with health care access and/or education, may present to the hospital later in the course of illness and, therefore, have longer, more complicated, and more costly hospital courses.

The effect of case mix on cost raises important questions about which estimates to use in cost-effectiveness models. It requires the analyst to make assumptions about the proportion of hospitalized patients who will have a high-risk condition and to estimate separately the cost of hospitalization for low- and high-risk patients. It is clear from our analysis that the cost of caring for a patient with a high-risk condition is potentially much greater than that estimated in most previous studies. Even among the low-risk patients in our study who were managed exclusive on the ward (n = 323; median length of stay: 2 days), the total mean hospitalization cost was $5687 (IQR: $3204–$6329), which is closer to the upper limit estimated by Meltzer et al but still nearly twice the cost reported in other previous studies. We conclude, then, that most previous studies of hospitalization costs for children with influenza may have systematically underestimated the true cost of influenza-related hospitalizations by underestimating the proportion of patients with high-risk conditions and/or the cost of hospitalization for patients with and without high-risk conditions. At best, previous estimates may reflect the cost of uncomplicated influenza-related hospitalizations in low-risk children, but they do not seem to represent the cost of hospitalization for children in the target group for vaccination according to ACIP criteria or those hospitalized in the ICU.

	POLICY IMPLICATIONS

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 
Our higher cost estimates for influenza-related hospitalization may impact the results of economic evaluations of influenza vaccination. Although lack of information about antecedent influenza vaccination prevented us from determining the proportion of hospitalization costs attributable to missed vaccination opportunities, recent cost-effectiveness analyses suggest that, depending on risk factors and vaccine cost, annual influenza vaccination is either cost-saving or relatively cost-effective in children.^8,9,20,21 In these analyses, lost parental productivity and outpatient expenses usually contribute more than hospitalization costs to the overall cost-effectiveness equation. However, doubling the baseline estimate for inpatient costs can only make annual influenza vaccination more cost-effective or cost-saving. This is especially true among children with high-risk characteristics, for whom the mean total inpatient cost in our study was approximately $15000 per patient. It is probable that in high-risk children, the inpatient cost estimates reported here would actually exert more leverage on net costs than lost parental productivity and outpatient expenses. Future cost-effectiveness analyses must incorporate these higher inpatient cost estimates.

Our data also provide additional economic evidence to support the ACIP recommendation that children with certain chronic medical conditions receive annual influenza vaccine. With the exception of immunosuppressive disorders, hemoglobinopathies, and chronic pulmonary disease, all other ACIP-designated high-risk medical conditions were associated with hospitalization costs >15th percentile ($15000). Ensuring that children with these conditions receive annual influenza vaccination could result in significant cost savings.

	CONCLUSIONS

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 
Previous economic evaluations have underestimated the cost of influenza-related hospitalizations in children. The presence of certain ACIP-designated chronic medical conditions is associated with higher influenza-related hospitalization costs. Immunizing patients with these conditions has the potential for significant cost savings.

	ACKNOWLEDGMENTS

 
This work was supported by grant H23/CCH32253-02 from the Centers for Disease Control and Prevention. Dr Keren was supported by grant K23 HD043179 from the National Institute of Child Health and Human Development, Bethesda, MD.

	FOOTNOTES

 
Accepted May 24, 2006.

Address correspondence to Ron Keren, MD, MPH, Children's Hospital of Philadelphia, 3535 Market St, Room 1524, Philadelphia, PA 19104. E-mail: keren{at}email.chop.edu

The authors have indicated they have no financial relationships relevant to this article to disclose.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION POLICY IMPLICATIONS CONCLUSIONS REFERENCES

 

1. Thompson WW, Shay DK, Weintraub E, et al. Influenza-associated hospitalizations in the United States. JAMA. 2004;292 :1333 –1340[Abstract/Free Full Text]
2. Thompson WW, Shay DK, Weintraub E, et al. Mortality associated with influenza and respiratory syncytial virus in the United States. JAMA. 2003;289 :179 –186[Abstract/Free Full Text]
3. Barker WH, Mullooly JP. Impact of epidemic type A influenza in a defined adult population. Am J Epidemiol. 1980;112 :798 –811[Abstract/Free Full Text]
4. Izurieta HS, Thompson WW, Kramarz P, et al. Influenza and the rates of hospitalization for respiratory disease among infants and young children. N Engl J Med. 2000;342 :232 –239[Abstract/Free Full Text]
5. Neuzil KM, Mellen BG, Wright PF, Mitchel EF Jr, Griffin MR. The effect of influenza on hospitalizations, outpatient visits, and courses of antibiotics in children. N Engl J Med. 2000;342 :225 –231[Abstract/Free Full Text]
6. Neuzil KM, Wright PF, Mitchel EF Jr, Griffin MR. The burden of influenza illness in children with asthma and other chronic medical conditions. J Pediatr. 2000;137 :856 –864[CrossRef][ISI][Medline]
7. Simonsen L, Fukuda K, Schonberger LB, Cox NJ. The impact of influenza epidemics on hospitalizations. J Infect Dis. 2000;181 :831 –837[CrossRef][ISI][Medline]
8. Cohen GM, Nettleman MD. Economic impact of influenza vaccination in preschool children. Pediatrics. 2000;106 :973 –976[Abstract/Free Full Text]
9. Luce BR, Zangwill KM, Palmer CS, et al. Cost-effectiveness analysis of an intranasal influenza vaccine for the prevention of influenza in healthy children. Pediatrics. 2001;108(2). Available at: www.pediatrics.org/cgi/content/full/108/2/e24
10. Meltzer MI, Cox NJ, Fukuda K. The economic impact of pandemic influenza in the United States: priorities for intervention. Emerg Infect Dis. 1999;5 :659 –671[ISI][Medline]
11. Keren R, Zaoutis TE, Bridges CB, et al. Neurological and neuromuscular disease as a risk factor for respiratory failure in children hospitalized with influenza infection. JAMA. 2005;294 :2188 –2194[Abstract/Free Full Text]
12. Centers for Disease Control and Prevention. Prevention and control of influenza: recommendations of the Advisory Committee on Immunization Practices (ACIP) [published correction appears in MMWR Morb Mortal Wkly Rep. 2005;54:750]. MMWR Recomm Rep. 2005;54 (RR-8):1–40
13. Centers for Medicare and Medicaid Services. PFS relative value files. Available at: www.cms.hhs.gov/PhysicianFeeSched/PFSRVF/list.asp#TopOfPage. Accessed February 27, 2006
14. US Department of Labor, Bureau of Labor Statistics. Consumer price indexes. Available at: www.bls.gov/cpi/home.htm. Accessed February 27, 2006
15. Zhou XH, Gao S, Hui SL. Methods for comparing the means of two independent log-normal samples [published correction appears in Biometrics. 1997;53:1566]. Biometrics. 1997;53 :1129 –1135[CrossRef][ISI][Medline]
16. Zhou XH, Melfi CA, Hui SL. Methods for comparison of cost data. Ann Intern Med. 1997;127 :752 –756[Abstract/Free Full Text]
17. Centers for Disease Control and Prevention, National Center for Health Statistics. National hospital discharge survey. Available at: www.cdc.gov/nchs/about/major/hdasd/nhdsdes.htm. Accessed on February 27, 2006
18. Stanford R, McLaughlin T, Okamoto LJ. The cost of asthma in the emergency department and hospital. Am J Respir Crit Care Med. 1999;160 :211 –215[Abstract/Free Full Text]
19. Hall JL, Katz BZ. Cost of influenza hospitalization at a tertiary care children's hospital and its impact on the cost-benefit analysis of the recommendation for universal influenza immunization in children age 6 to 23 months. J Pediatr. 2005;147 :807 –811[CrossRef][ISI][Medline]
20. Meltzer MI, Neuzil KM, Griffin MR, Fukuda K. An economic analysis of annual influenza vaccination of children. Vaccine. 2005;23 :1004 –1014[CrossRef][ISI][Medline]
21. White T, Lavoie S, Nettleman MD. Potential cost savings attributable to influenza vaccination of school-aged children. Pediatrics. 1999;103(6) . Available at: www.pediatrics.org/cgi/content/full/103/6/e73

This Article Abstract Full Text (PDF) An erratum has been published P3Rs: Submit a response Alert me when this article is cited Alert me when P3Rs are posted Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Keren, R. Articles by Coffin, S. E. Search for Related Content PubMed PubMed Citation Articles by Keren, R. Articles by Coffin, S. E. Related Collections Infectious Disease & Immunity

	Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved.